Comparing Scaled Agile Framework (SAFe) and Disciplined Agile Delivery (DAD)

Editor's Notes

• #2: Session Title:  Comparing Scaled Agile Approaches:  Scaled Agile Framework (SAFe) and Disciplined Agile Delivery (DAD)Session Description:  Leveraging established and proven Scrum, XP, and lean software development methods, agile scaling frameworks such as Dean Leffingwell's Scaled Agile Framework (SAFe) and Scott Ambler's Disciplined Agile Delivery (DAD) provide agile process frameworks for addressing large, complex software delivery initiatives through their full delivery lifecycle from project initiation to production.   These frameworks have received significant interest in both federal government and private industries, recognizing the need for team-based iterative and incremental adaptive approaches to software development, balanced with scaling processes and factors at the Program and Portfolio levels and the for an organizational governance model and guidance for large enterprise engagements.  This session will provide a brief overview of these two agile scaling models, address the benefits of what both are trying to accomplish, and compare and contrast specific similarities and differences.   Track:  Government or Improving your GameStyle:  LectureDuration:  60 minutesBoth DAD and SAFe are Hybrid Process frameworks - detailing how to scale agile to the enterpriseScaled Agile Framework (SAFe) since; Scaledagileframework.comDisciplined Agile Delivery (DAD); Disciplinedagiledelivery.comCertifications and Industry AwarenessSAFe SA and SPC certificationsDAD certifications - Yellow, Green, Black Belt http://disciplinedagileconsortium.org/
• #3: Talking Points:We assume you have an understanding of agile, scrum, xp foundations (e.g. roles, teams, activities/events, artifacts)...this presentationis about agile scalingRoles: SM, PO, Self Organizing TeamEvents: Planning, Daily standup, Retro, DemoArtifacts: Backlog, DoD, Burndown metrics
• #4: Talking Points:Review / highlight some Key Characteristics of Agile Delivery to establish baseline assumptions of agile benefits to agile deliveryEx: Iterative and incremental, Just-enough elaboration of requirements (user stories), Transparency, Code quality, Automated testingBoth scaling frameworks have Lean and Agile Underpinnings/FoundationsScrumKanbanXP/Agile EngineeringAgile Modeling (DAD)Agile Data (DAD)Lean Software Development - Product Development Flow (SAFe)Agile Architecture (SAFe and DAD)Principles and Values - Agile CultureAgile Manifesto & Principles (DAD and SAFe)Transparency, Program Execution, Code Quality, Alignment (SAFe)
• #5: Talking Points:A typical agile teamHighlight automated test engineer
• #6: Talking Points: Reference agile scaling supporting a dozen or more agile delivery teams with supporting governance constructs (supporting and service roles, teams, etc.) that allows you to execute agile-at-scaleMention Ch33 experience within the Federal Government ecosystem300 person programDozens of Scrum Delivery and Service teamsPre-SAFe and DAD
• #7: Talking Points: A large-scale Agile enterprise approach – PARTICULARLY IN THE FEDERAL GOVERNMENT SPACE
• #8: Talking Points:The “Big Picture” graphic  highlights the individual roles, teams, activities/events and artifacts necessary to scale agile from the team to program to the enterprise level - emphasizes continuous flowSAFe is about scaling software development for our programs and taking an enterprise approachNotice this framework can support multiple Programs under one Portfolio and multipleTeams under one ProgramThere are many icons across the Big Picture that represent people within the process framework Tiered governance results in timely decisions, accountability, and traceability Active collaboration ensures traceability, alignment to vision and continuous discovery of requirements Emphasis on Code Quality through eXtreme Programming (XP) to foster better code Daily builds and production-ready releases within a 3-month period Efficiency gained by a fully automated, Agile testing process that is consistent and repeatable Real-time dashboards provide valuable insight to all stakeholdersOther Points: Frequent backlog grooming lowers cost and yields higher quality code Scalability, responsiveness, and standards drive design Powerful, virtualized CI “ecosystem” provides faster time to market and higher quality code CI ecosystem automates many Configuration Management (CM) tasks in a repeatable, consistent fashion Experienced Information Assurance (IA) leadership knowledgeable on VA-specific security requirementsSAFe provides a formal 3-tiered governance structure with supporting processes torealize benefits of agile at the Team, Program, and Portfolio levels ANDmitigate risks associated with scaling and coordination across project teams and stakeholders with other VistA enterprise efforts. Portfolio Vision, Program Release management, and Team Sprint planningWe will deliver timely decision-making by following agile value and change management practices, prioritizing issues and requirements through regular reviews and retrospection, and resolving impediments at the appropriate tier. Access to Frequent, Transparent, Real-time Information—We will provide relevant program dashboards with real-time backlog delivery metrics, enabling insight into actual progress against objectives. Our approach ensures that customers actively set direction during project initiation and regularly refine the vision at subsequent release and Sprint start points.
• #9: Talking Points:Talk to Flow of Work into Portfolio Level – Inputs and Outputs including investment themes drive business and architectural Program Epics and Vision Highlight key Portfolio Level Roles, Events, Artifacts, Inputs/Outputs:Specific Portfolio level roles, events, and artifacts exist to support agile-at-scale – Portfolio Management Team, Enterprise ArchitectEstablish a Portfolio Management Team composed of business and architecture organizational leadership to prioritize high-level Epic requirements that map to investment prioritiesVision guides investment priorities and enterprise architecture Kanban system provides a structured portfolio management approach to iteratively evolve Portfolio Backlog priorities.We will use a Kanban process to focus the high-level stakeholder ideas and requirements around the most impactful areas defined as a resulting set of Epics. Portfolio Backlog includes both functional and architectural epics and span multiple project Releases – enterprise architecture is a “first class citizen”.We will incorporate governance and direction from enterprise-level organizationsto ensure alignment to the organizationally-defined investment priorities. Objective, real-time portfoliometrics& dashboard support governance and continuous improvement============SAFe Guidelines: Lean principles emphasize sustainably fastest value delivery Portfolio vision guides investments and the enterprise architecture needed to support customer and business needsBusiness epic kanban system provides visibility and work-in-process limits to support continuous product development flowEnterprise architecture is a first class citizen; architectural epics are developed and maintained in a kanban systemObjective metrics (at all levels) support governance and kaizenStructured approach clarifying and approving Portfolio Backlog with EpicsClarifies additional Portfolio level roles, agile practices/techniques, teams needed for supporting agile-at-scale
• #10: Talking Points:Talk to Flow of Work into Program Level – inputs and outputs and how Portfolio Vision drives Program Backlog, Product Roadmap and Product VisionHighlight what some acronyms meanHighlight key Program Level Roles, Events, Artifacts, Inputs/Outputs:System Level: Planning, Test, Integration, Demo, GovernanceProgram / PSI-Release focus supports the delivery of fully tested, full-featured solutionsevery 3 months Face-to-face PSI-Release planning cadence provides development collaboration, alignment, synchronization, evaluationProduct / ProgramManagement :Decomposes Epics into business and technical Features and regularly grooms the Program Backlog; consumed by Scrum Team (POs)Participates in System Demos and approves Feature increments every 2 weeksDevelops Product Roadmap and Vision lists desired functionality for up to 3 future PSI-Releases summarized from the prioritized Program Backlog FeaturesDevOps will manage and extend an automated CI ecosystem and self-service infrastructureSystem Team supports system integration,cross-team E2E testingactivities, and System DemosSystem Architect and UX defines the technical elements and standards across all Program Release deliverablesRelease Management and Defect Management roles with key cross-team coordination and collaboration rolesEmbedded UAT Team required for every PSI/Release to incrementally plan and test Feature solutionsAggregated, real-time program metrics & dashboardInspect & Adapt Workshopat end of every PSI-Release boundary==============SAFe Guidelines:A self-organizing, self-managing team-of-agile-teams committed to continuous value delivery, aligned to a common missionDelivers fully tested, system-level solutions every 8-12 weeksCommon sprint lengths and normalized velocityRegular Architecture alignment via the Architectural Features and RunwayFace-to-face planning cadence provides development collaboration, alignment, synchronization, evaluation through System DemoReoccurring embedded UAT approachClarifies additional Program level roles, agile practices/techniques, teams needed for supporting agile-at-scale
• #11: Talking Points:Part of the SAFe framework most are probably already familiar with…Scrum, XP, etc.Talk to Flow of Work into Team Level – inputs and outputs and how Features drive and how things work highlighting new conceptsHighlight key Team Level Roles, Events, Artifacts, Inputs/Outputs:Empowered self-managing Teams that deliver fully tested, working software (user stories) every 2 weeksApply Scrum project management and eXtreme Programming (XP) technical approaches and practices (e.g. Sprint Planning, Daily Stand Ups, Sprint Demo, etc.)Dedicated Product Owner on each 7-9 member cross-functional Team – each with unique Domain focusThe Scrum Master serves as an Agile Coach for multiple TeamsTeams submit their component outputs to the program-level System Team for daily integration into the baselineTeams operate under common Vision linked through Epics, Features, StoriesBuilt in Hardening, Innovation, and Planning (HIP) Sprint supports needed refactoring and future event planningReal-time team metrics & dashboard support continuous improvement
• #12: Talking Points:DAD founders recognized that mainstream agile does NOT provide enough guidance for typical enterprises – mature implementations recognize an additional level of rigor, such as governance, architecture planning, and modelingDAD takes a goal-driven approachthat describes agile implementation and tailoring strategies to meet the needs of your organization’s situationToolkit: Adopts, extends, tailors mainstream agile practices, strategies, patterns from a variety of sourcesDAD Agile Scaling Model (ASM)defines a roadmap to effectivelyadopt and tailor agile strategies by: Foundation: leverage core agile methods Framework: adopting a disciplined agile delivery lifecycle Factor: recognize agile Scaling Factors to tailor adopted strategies to address delivery team complexities
• #13: Talking Points:End-to-end, full agile delivery lifecycle thatincludes:Inception, Construction, and Transition phases from beginning of project to production Each phase reflects agile Coordinate-Collaborate-Conclude (3C) rhythmLightweight explicit “milestones” and events to ensure program / teams focused on right things – not just construction aspects – governance and risk reduction strategyGoals driven approach = each agile delivery phase has a set of goals with correspondingcommon strategies and guidance for solution/delivery teams to tailor the process to address the context of their situationHybrid in that it extends Scrum, XP, and other proven agile methods to includes advice about technical practices (missing from Scrum) as well as modeling, documentation, and governance strategiesGoldilocks, just enough process to limit overly prescriptive or to tailor up processes that prevent reinventing / relearning techniques that Scrum/XP does not ocverProvides advice, alternatives and trade-offs in agile adoption strategy in a structured and disciplined wayOverall DAD delivery goals to fulfill project mission, grow team member skills, enhance existing infrastructure, improve team processes, address risks
• #14: Talking Points:2 levels of Inception:During one-time Project Initiation event at Program Level to get things going – like "Iteration 0" project readiness activities including On-boarding and trainingDuring iterative Release Initiation events for each PSI-ReleaseInception Phase goalhighlights:Forming and on-boarding of teamSet up work environment Get enough infrastructure to get startedIdentify high level business objectives for the project and perform lightweight visioning activities to properly frame the projectPerform architecture envisioning to ensure design going in the right direction Perform PSI-Release PlanningIdentify risks and address other PMP component
• #15: Talking Points:2 Paths to Construction Phase: Iterative Solution delivery - Emphasis on “standard” iterative working slice of solutionIntegrated Solution delivery – Emphasis on end to end working slice of solutionConstruction Phase goalhighlights:Period of time where required functionality is builtwithin timeboxed iterationsEnd of each iteration a demonstrable increment of potentially consumable solution has been produced and regression testedDelivery team applies a hybrid of “Standard” practices from Scrum, XP, Agile Modeling, Agile Data, and other methods to deliver the solutionSupports Daily Stand-up coordinationSupports Iteration planning and modelingBacklogsIteration DemoRetrosSpikesAgile metricsCIStrong emphasis on “Advanced” agile engineering practicesTDD/ATDDCDLook-Ahead planning and modelingContinuous documentationReleases determined by comparing whether there is sufficient functionality to justify the cost of transition – minimally marketable release (MMR)Performsome look-ahead planning and modeling, often referred to as backlog grooming, on a just-in-time basis to be more effective when implementing work items during each iterationLeverage the existing organizational ecosystem to reuse existing services, patterns, templates, code, guidelines and other assetsEncourage project specific process improvement through self-organization within the constraints that make sense for the organizationAdapt the types and formality of work products produced for the context of the projectsDoes not say who (person and team) performs action within approach like SAFe
• #16: Talking Points:2 Paths to Construction Phase: Iterative Solution delivery - Emphasis on “standard” iterative working slice of solutionIntegrated Solution delivery – Emphasis on end to end working slice of solutionConstruction Phase goalhighlights:Period of time where required functionality is builtwithin timeboxed iterationsEnd of each iteration a demonstrable increment of potentially consumable solution has been produced and regression testedDelivery team applies a hybrid of “Standard” practices from Scrum, XP, Agile Modeling, Agile Data, and other methods to deliver the solutionSupports Daily Stand-up coordinationSupports Iteration planning and modelingBacklogsIteration DemoRetrosSpikesAgile metricsCIStrong emphasis on “Advanced” agile engineering practicesTDD/ATDDCDLook-Ahead planning and modelingContinuous documentationReleases determined by comparing whether there is sufficient functionality to justify the cost of transition – minimally marketable release (MMR)Performsome look-ahead planning and modeling, often referred to as backlog grooming, on a just-in-time basis to be more effective when implementing work items during each iterationLeverage the existing organizational ecosystem to reuse existing services, patterns, templates, code, guidelines and other assetsEncourage project specific process improvement through self-organization within the constraints that make sense for the organizationAdapt the types and formality of work products produced for the context of the projectsDoes not say who (person and team) performs action within approach like SAFe
• #17: Talking Points:2 levels of Transition:During one-time Project Transition event at Program Level to address end of lifecycle testing and readiness before going into ProductionDuring iterative Release Transition event for each PSI-Release -- like “HIP Sprint" production readiness activities for a ReleaseTransition Phase goalhighlights:Focus on delivering system into Production across one or more PSI-ReleasesStakeholder training and documentationProduction infrastructure readiness and hardening; securityPerformEnd of lifecycle testing (e.g. UAT, performance, etc.)DAD recognizes that for sophisticated enterprise agile projects deploying the solution to the stakeholders is often not a trivial exerciseDAD teams, as well as the enterprise overall, will streamline their deployment processes so that over time this phase become shorters and ideally disappears from adopting continuous deployment strategiesIt takes discipline to evolve Transition from a phase to an activity.
• #18: Talking Points: GOVERNANCEEncourages collaboration with corporate stakeholders such as Enterprise Architects, Portfolio Managers, Program Management, DevOps, IT InfrastructureGovernance issue that address corporate assets: team roles and responsibilities, governing bodies, exceptions and escalation processes, knowledge sharing processes, metrics strategy, risk mitigation, reward structure, status reporting, audit processes, policies/standards/guidelines, artifacts and their lifecycles, etc.SCALING FACTORSDAD process framework provides a scalable foundation for agile deliveriesDAD requires pragmatic adaption of methods and approaches in context with environment and constraintsLarge-scale agile projects range of problem and solution complexities, so scaling factors need to be consider when tailoring your agile strategy, such as: Team size. Agile teams may range from as small as two people to hundreds and potentially thousands of people.Geographical distribution. A team may be located in a single room, on the same floor but in different offices or cubes, in the same building, in the same city, or even in different cities around the globe.Organizational distribution. Some agile teams are comprised of people who work for the same group in the same company. Other teams have people from different groups of the same company. Some teams are made up of people from similar organizations working together as a consortium. Some team members may be consultants or contractors. Sometimes some of the work is outsourced to one or more external service provider(s).Regulatory compliance. Some agile teams must conform to industry regulations such as the Dodd-Frank act, Sarbanes-Oxley, or Food and Drug Administration (FDA) regulations.Domain complexity. Some teams apply agile techniques in straightforward situations, such as building an informational Web site, to more complex situations such as building an internal business application, and even in life-critical health-care systems.Technical complexity. Some agile teams build brand-new, “greenfield systems” from scratch running on a single technology platform with no need to integrate with other systems. At the other end of the spectrum some agile teams are working with multiple technologies, evolving and integrating with legacy systems, and evolving and accessing legacy data sources.Organizational complexity. In some organizations people work to the same vision and collaborate effectively. Other organizations suffer from politics. Some organizations have competing visions for how people should work and worse yet have various subgroups following and promoting those visions.Enterprise discipline. Many organizations want their teams to work toward a common enterprise architecture, take advantage of strategic reuse opportunities, and reflect their overall portfolio strategy.
• #19: Talking Points: Documentation does not go away particularly on FedGovt programsLightweight approaches congruent/tailored to fit within frameworkTechniques include combine, living documentation, agile artifactsExamples include Tangent PMP and metrics
• #39: Talking Points: Leverage these frameworks and make them work within the context of your environmentBoth DAD and SAFe are Hybrid Process frameworks - detailing how to scale agile to the enterpriseScaled Agile Framework (SAFe) since; Scaledagileframework.comDisciplined Agile Delivery (DAD); Disciplinedagiledelivery.comCertifications and Industry AwarenessSAFe SA and SPC certificationsDAD certifications - Yellow, Green, Black Belt http://disciplinedagileconsortium.org/
• #40: Talking Points:Leverage these frameworks and make them work within the context of your environmentTechnical architecture & development focus includes XP, agile modeling, agile data, and lean software development. Both DAD and SAFe are Hybrid Process frameworks - detailing how to scale agile to the enterpriseScaled Agile Framework (SAFe) since; Scaledagileframework.comDisciplined Agile Delivery (DAD); Disciplinedagiledelivery.comCertifications and Industry AwarenessSAFe SA and SPC certificationsDAD certifications - Yellow, Green, Black Belt http://disciplinedagileconsortium.org/